Vision-CAIR / MiniGPT-4

Open-sourced codes for MiniGPT-4 and MiniGPT-v2 (https://minigpt-4.github.io, https://minigpt-v2.github.io/)
https://minigpt-4.github.io
BSD 3-Clause "New" or "Revised" License
25.4k stars 2.91k forks source link

The same input but different output from the online demo. #487

Open changbaozhou opened 8 months ago

changbaozhou commented 8 months ago

I deployed minigpt-4 according to huggingface demo on my server, when I input the same image with the same prompt "Describe this image". to my local program, I got the different output with the online demo. Could anyone tell me the reason?

image

and the following is the output of my demo: This image shows a group of people gathered around a woman who is being hugged by another person. The woman is wearing a pink sari and has her arms wrapped around the person hugging her. The other people in the image are also hugging and embracing each other. The image appears to be taken in a public place, such as a park or street, and the people are smiling and laughing. The overall atmosphere of the image is one of joy and celebration.

It seems that the output of online demo is more accurate than mine.

changbaozhou commented 8 months ago

image this is my config.yaml